diff --git a/NEWS b/NEWS
index 0f941190e12c0d9d50ac49b3c90cee4659e071ff..cc3d07157a9985e01be09dfa36530b7a2532d519 100644
--- a/NEWS
+++ b/NEWS
@@ -81,6 +81,9 @@ documents those changes that are of interest to users and administrators.
-- Functions slurm_load_node() slurm_load_partitions() modified to show all
nodes/partitions in a federation when the SHOW_GLOBAL flag is used.
+* Changes in Slurm 17.02.4
+==========================
+
* Changes in Slurm 17.02.3
==========================
-- Increase --cpu_bind and --mem_bind field length limits.
@@ -137,6 +140,14 @@ documents those changes that are of interest to users and administrators.
-- MYSQL - Fix memory leak when loading archived jobs into the database.
-- Fix potential race condition when starting the priority/multifactor plugin's
decay thread.
+ -- Sanity check to make sure we have started a job in acct_policy.c before we
+ clear it as started.
+ -- Allow reboot program to use arguments.
+ -- Message Aggr - Remove race condition on slurmd shutdown with respects to
+ destroying a mutex.
+ -- Fix updating job priority on multiple partitions to be correct.
+ -- Don't remove admin comment when updating a job.
+ -- Return error when bad separator is given for scontrol update job licenses.
* Changes in Slurm 17.02.2
==========================
diff --git a/src/common/msg_aggr.c b/src/common/msg_aggr.c
index 9d371b632bd2659e5681104e1d3d816853eda9ca..5ab41f17c1d5f441a0aaca9117777c9506f63323 100644
--- a/src/common/msg_aggr.c
+++ b/src/common/msg_aggr.c
@@ -333,7 +333,6 @@ extern void msg_aggr_sender_fini(void)
FREE_NULL_LIST(msg_collection.msg_aggr_list);
slurm_mutex_unlock(&msg_collection.aggr_mutex);
FREE_NULL_LIST(msg_collection.msg_list);
- slurm_mutex_destroy(&msg_collection.aggr_mutex);
slurm_mutex_destroy(&msg_collection.mutex);
}
@@ -342,6 +341,7 @@ extern void msg_aggr_add_msg(slurm_msg_t *msg, bool wait,
{
int count;
static uint16_t msg_index = 1;
+ static uint32_t wait_count = 0;
if (!msg_collection.running)
return;
@@ -388,12 +388,16 @@ extern void msg_aggr_add_msg(slurm_msg_t *msg, bool wait,
timeout.tv_sec = now.tv_sec + msg_timeout;
timeout.tv_nsec = now.tv_usec * 1000;
+ wait_count++;
if (pthread_cond_timedwait(&msg_aggr->wait_cond,
&msg_collection.aggr_mutex,
&timeout) == ETIMEDOUT)
_handle_msg_aggr_ret(msg_aggr->msg_index, 1);
+ wait_count--;
slurm_mutex_unlock(&msg_collection.aggr_mutex);
-
+;
+ if (!msg_collection.running && !wait_count)
+ slurm_mutex_destroy(&msg_collection.aggr_mutex);
_msg_aggr_free(msg_aggr);
}
diff --git a/src/common/slurm_selecttype_info.c b/src/common/slurm_selecttype_info.c
index b38b837e875be6ee580e5c54d6053f2ca21f326b..9c50e812247cfe0cddfa06b2c4b0b8068a10ed39 100644
--- a/src/common/slurm_selecttype_info.c
+++ b/src/common/slurm_selecttype_info.c
@@ -52,7 +52,6 @@ int parse_select_type_param(char *select_type_parameters, uint16_t *param)
int rc = SLURM_SUCCESS;
char *str_parameters, *st_str = NULL;
int param_cnt = 0;
- bool one_task = false;
*param = 0;
st_str = xstrdup(select_type_parameters);
@@ -82,9 +81,6 @@ int parse_select_type_param(char *select_type_parameters, uint16_t *param)
*param |= CR_CPU;
*param |= CR_MEMORY;
param_cnt++;
- } else if (!xstrcasecmp(str_parameters,
- "CR_ONE_TASK_PER_CORE")) {
- one_task = true;
} else if (!xstrcasecmp(str_parameters, "other_cons_res")) {
*param |= CR_OTHER_CONS_RES;
} else if (!xstrcasecmp(str_parameters,
@@ -119,7 +115,7 @@ int parse_select_type_param(char *select_type_parameters, uint16_t *param)
return rc;
}
- if ((*param & CR_CPU) && one_task) {
+ if ((*param & CR_CPU) && (*param & CR_ONE_TASK_PER_CORE)) {
error("CR_ONE_TASK_PER_CORE is not compatible with CR_CPU*, please change to use CR_CORE* instead.");
rc = SLURM_ERROR;
xfree(st_str);
diff --git a/src/slurmctld/acct_policy.c b/src/slurmctld/acct_policy.c
index a89384d390f607331ca85482ae121ac3dbcb6174..e3f5c57897462fde669000c2fc1e01bb23082b2b 100644
--- a/src/slurmctld/acct_policy.c
+++ b/src/slurmctld/acct_policy.c
@@ -546,6 +546,13 @@ static void _qos_adjust_limit_usage(int type, struct job_record *job_ptr,
used_limits_a->jobs++;
break;
case ACCT_POLICY_JOB_FINI:
+ /*
+ * If tres_alloc_cnt doesn't exist means ACCT_POLICY_JOB_BEGIN
+ * was never called so no need to clean up that which was never
+ * set up.
+ */
+ if (!job_ptr->tres_alloc_cnt)
+ break;
qos_ptr->usage->grp_used_jobs--;
if ((int32_t)qos_ptr->usage->grp_used_jobs < 0) {
qos_ptr->usage->grp_used_jobs = 0;
diff --git a/src/slurmctld/job_mgr.c b/src/slurmctld/job_mgr.c
index 8fe016a86f21b86d864febc2a6726fb525e1aed1..a4a1faa234ef5913de379959d1c40cfdcbe8c627 100644
--- a/src/slurmctld/job_mgr.c
+++ b/src/slurmctld/job_mgr.c
@@ -10960,23 +10960,29 @@ static int _update_job(struct job_record *job_ptr, job_desc_msg_t * job_specs,
}
/* Always do this last just in case the assoc_ptr changed */
- if (job_specs->admin_comment && !validate_super_user(uid)) {
- error("Attempt to change admin_comment for job %u",
- job_ptr->job_id);
- error_code = ESLURM_ACCESS_DENIED;
- } else if (job_specs->admin_comment &&
- (job_specs->admin_comment[0] == '+') &&
- (job_specs->admin_comment[1] == '=')) {
- if (job_ptr->admin_comment)
- xstrcat(job_ptr->admin_comment, ",");
- xstrcat(job_ptr->admin_comment, job_specs->admin_comment + 2);
- info("update_job: setting admin_comment to %s for job_id %u",
- job_ptr->admin_comment, job_ptr->job_id);
- } else {
- xfree(job_ptr->admin_comment);
- job_ptr->admin_comment = xstrdup(job_specs->admin_comment);
- info("update_job: setting admin_comment to %s for job_id %u",
- job_ptr->admin_comment, job_ptr->job_id);
+ if (job_specs->admin_comment) {
+ if (!validate_super_user(uid)) {
+ error("Attempt to change admin_comment for job %u",
+ job_ptr->job_id);
+ error_code = ESLURM_ACCESS_DENIED;
+ } else if ((job_specs->admin_comment[0] == '+') &&
+ (job_specs->admin_comment[1] == '=')) {
+ if (job_ptr->admin_comment)
+ xstrcat(job_ptr->admin_comment, ",");
+ xstrcat(job_ptr->admin_comment,
+ job_specs->admin_comment + 2);
+ info("update_job: adding to admin_comment it is now %s for job_id %u",
+ job_ptr->admin_comment, job_ptr->job_id);
+ } else if (!xstrcmp(job_ptr->admin_comment,
+ job_specs->admin_comment)) {
+ info("update_job: admin_comment the same as before, not changing");
+ } else {
+ xfree(job_ptr->admin_comment);
+ job_ptr->admin_comment =
+ xstrdup(job_specs->admin_comment);
+ info("update_job: setting admin_comment to %s for job_id %u",
+ job_ptr->admin_comment, job_ptr->job_id);
+ }
}
/* Always do this last just in case the assoc_ptr changed */
@@ -11472,6 +11478,15 @@ static int _update_job(struct job_record *job_ptr, job_desc_msg_t * job_specs,
} else
error_code = ESLURM_PRIO_RESET_FAIL;
job_ptr->priority = job_specs->priority;
+ if (job_ptr->part_ptr_list &&
+ job_ptr->priority_array) {
+ int i, j = list_count(
+ job_ptr->part_ptr_list);
+ for (i = 0; i < j; i++) {
+ job_ptr->priority_array[i] =
+ job_specs->priority;
+ }
+ }
}
info("sched: update_job: setting priority to %u for "
"job_id %u", job_ptr->priority,
diff --git a/src/slurmctld/licenses.c b/src/slurmctld/licenses.c
index 0e65617acf74ad03334f6a56ddcd4732759645d6..2a7053e3f580190e09bf59fe337f37e2a55b2b6b 100644
--- a/src/slurmctld/licenses.c
+++ b/src/slurmctld/licenses.c
@@ -148,9 +148,12 @@ static List _build_license_list(char *licenses, bool *valid)
if ((token[i] == ':') || (token[i] == '*')) {
token[i++] = '\0';
num = (int32_t)strtol(&token[i], &end_num, 10);
+ if (*end_num != '\0')
+ *valid = false;
+ break;
}
}
- if (num < 0) {
+ if (num < 0 || !(*valid)) {
*valid = false;
break;
}
diff --git a/src/slurmctld/node_scheduler.c b/src/slurmctld/node_scheduler.c
index 9f27685201073e5c724ab3a1e604c06507a0a33b..8944e75c41f84b55a50192faee7b9be2239baf2a 100644
--- a/src/slurmctld/node_scheduler.c
+++ b/src/slurmctld/node_scheduler.c
@@ -2511,10 +2511,6 @@ extern int select_nodes(struct job_record *job_ptr, bool test_only,
if (select_g_job_begin(job_ptr) != SLURM_SUCCESS) {
/* Leave job queued, something is hosed */
error("select_g_job_begin(%u): %m", job_ptr->job_id);
-
- /* Since we began the job earlier we have to cancel it */
- (void) bb_g_job_cancel(job_ptr);
-
error_code = ESLURM_NODES_BUSY;
job_ptr->start_time = 0;
job_ptr->time_last_active = 0;
@@ -2534,10 +2530,6 @@ extern int select_nodes(struct job_record *job_ptr, bool test_only,
error("Select plugin failed to set job resources, nodes");
/* Do not attempt to allocate the select_bitmap nodes since
* select plugin failed to set job resources */
-
- /* Since we began the bb job earlier we have to cancel it */
- (void) bb_g_job_cancel(job_ptr);
-
error_code = ESLURM_NODES_BUSY;
job_ptr->start_time = 0;
job_ptr->time_last_active = 0;
@@ -2557,11 +2549,6 @@ extern int select_nodes(struct job_record *job_ptr, bool test_only,
if (!job_ptr->job_resrcs) {
/* If we don't exit earlier the empty job_resrcs might
* be dereferenced later */
-
- /* Since we began the bb job earlier we have to cancel
- * it */
- (void) bb_g_job_cancel(job_ptr);
-
error_code = ESLURM_NODES_BUSY;
job_ptr->start_time = 0;
job_ptr->time_last_active = 0;
diff --git a/src/slurmd/slurmd/req.c b/src/slurmd/slurmd/req.c
index 907e4678db16bf7fd3ce9b1c505548bad47383a7..3f1e661123f98b80721e069852f80dceebb1675a 100644
--- a/src/slurmd/slurmd/req.c
+++ b/src/slurmd/slurmd/req.c
@@ -2567,6 +2567,10 @@ _rpc_reboot(slurm_msg_t *msg)
sp = xstrdup(reboot_program);
reboot_msg = (reboot_msg_t *) msg->data;
if (reboot_msg && reboot_msg->features) {
+ /*
+ * Run reboot_program with only arguments given
+ * in reboot_msg->features.
+ */
info("Node reboot request with features %s being processed",
reboot_msg->features);
(void) node_features_g_node_set(
@@ -2578,7 +2582,8 @@ _rpc_reboot(slurm_msg_t *msg)
cmd = xstrdup(sp);
}
} else {
- cmd = xstrdup(sp);
+ /* Run reboot_program verbatim */
+ cmd = xstrdup(reboot_program);
info("Node reboot request being processed");
}
if (access(sp, R_OK | X_OK) < 0)
diff --git a/testsuite/expect/test1.88 b/testsuite/expect/test1.88
index be9041c939ea5ccf155de6e24ab6a21c17f6ec74..01ecf52600b1497d38da6dd550b193c86ef5a21d 100755
--- a/testsuite/expect/test1.88
+++ b/testsuite/expect/test1.88
@@ -109,7 +109,7 @@ make_bash_script $file_in "
#
set timeout $max_job_delay
set no_start 0
-set sbatch_pid [spawn $sbatch -N1-3 -n6 --output=$file_out --error=$file_err -t1 $file_in]
+set sbatch_pid [spawn $sbatch -N1-6 -n6 --output=$file_out --error=$file_err -t1 $file_in]
expect {
-re "Submitted batch job ($number)" {
set job_id $expect_out(1,string)